Agent Skills
› lawve-ai/awesome-legal-skills
› law-expertise-to-skill-ignacio-adrian-lerer
law-expertise-to-skill-ignacio-adrian-lerer
GitHub将律师的专业知识、工作流和审查标准转化为安全、可检查的法律AI技能包。提取可复用能力并设定边界，生成结构化技能描述与使用说明，严禁冒充律师或提供法律建议，确保合规与人审要求。
Trigger Scenarios
用户希望将法律文书或经验转化为标准化AI技能
需要将法律审查标准或谈判策略封装为可复用模块
Install
npx skills add lawve-ai/awesome-legal-skills --skill law-expertise-to-skill-ignacio-adrian-lerer -g -y
SKILL.md
Frontmatter
{
"name": "law-expertise-to-skill-ignacio-adrian-lerer",
"metadata": {
"author": "Ignacio Adrián Lerer",
"license": "agpl-3.0",
"version": "2026-06-01"
},
"description": "Turn a lawyer's bounded expertise, workflow, review standard, or legal judgment pattern into a safe, inspectable legal AI skill. Use when a user provides legal memos, comments, checklists, negotiation notes, review habits, compliance playbooks, or professional feedback and wants a reusable Lawve-style skill. Do not impersonate the lawyer or claim to reproduce a person; distill only authorized, source-bounded capability."
}
Law Expertise To Skill
Use this skill to convert legal professional know-how into a simple, safe skill package.
Inputs
Accept:
- Legal workflow notes, memos, comments, checklists, templates, negotiation positions, audit criteria, or review standards.
- A short description of the target legal task.
- User feedback about what the skill should or should not do.
Do not use private client material, confidential firm know-how, privileged communications, or personal traces unless the user confirms authority to use them.
Workflow
- Define the legal capability: what the skill helps with, for whom, and under which legal system or practice area.
- Define source boundaries: what materials were used, what was excluded, and whether the evidence is first-hand, public, generated, or user-described.
- Extract only reusable capability:
- issue-spotting criteria;
- review sequence;
- risk taxonomy;
- drafting or negotiation heuristics;
- escalation thresholds;
- required disclaimers or source checks.
- Separate capability from voice:
- Capability is allowed.
- Professional tone guidance is allowed.
- Persona simulation or identity replacement is not allowed.
- Add safety boundaries:
- no legal advice beyond the skill scope;
- no unsupported jurisdictional claims;
- no hidden source inference;
- no confidential-data reuse;
- human lawyer review required for material decisions.
- Produce a compact skill with:
- clear trigger;
- negative scope;
- step-by-step method;
- output format;
- verification checklist.
Output
Return:
What this skill does: short marketplace description.How to use: user-facing instructions.SKILL.md: portable skill text.Source boundaries: what evidence was used and what was not.Human review points: decisions that need a lawyer.
Refusal / Escalation
Refuse or request clarification when the user asks to:
- impersonate a real lawyer without consent;
- use private traces from another person without authority;
- hide source limitations;
- remove human-review requirements from high-stakes legal work;
- present the skill as the actual person or as guaranteed legal advice.
Version History
- 8e51264 Current 2026-07-31 02:39
